Thermochemistry of carbonyl addition reactions. I. Addition of water

Sep 1, 1971 - Poly(arylene ether sulfone) Statistical Copolymers Bearing Perfluoroalkylsulfonic Acid Moieties. Haibo Li , Andrew B. Jackson , Nathan J...
0 downloads 0 Views 264KB Size
4596

Both I11 and IV exhibit only a sharp doublet for the methoxy protons, and for IV this requires that the phosphite ligands adopt the unambiguous pseudocis configuration indicated, with essentially zero P-P coupling. This also provides the basis for the simplifying assumptions in the interpretation of the spectrum of VI, discussed above. In the spectrum for V, Figure 1, we see a virtual triplet whose central band is of lower relative intensity than that of VI and whose maximum is displaced somewhat to lower field (T 6.22) from the midpoint (7 6.23) of the two outer peaks. This is the result of near coincidence in what we assign as a doublet for the phosphite (L') trans to CO and a virtual triplet of the two phosphites trans to each other (L and L"). The near superposition of an outer 1 : 1 doublet and a virtual 1 :2 : 1 triplet requires that the apparent triplet have peaks of relative intensity 1 :1.33: 1, close to that observed. The most prominent feature to be noted in the nmr spectra of 111-VI is the single chemical shift for metalhydrogen in all the derivatives. Each hydrogen appears also to be equivalently spin coupled to all the phosphorus nuclei in any particular complex. Whatever else may be said about the structures of these substituted clusters, it is clear that some form of intramolecular tautomerism is occurring in which the chemical shift of hydrogen and its coupling to phosphorus becomes averaged. The increase in averaged coupling constant from I11 through VI is interpretable as due to an increase in average proximity of H to P as ligand substitution is increased. The observed equivalence of the environment of the hydrogen atoms could come about from their rapid exchange of positions along the edges and in part from an intramolecular rotation of the RU(CO)~L groups about the C3 axes of the tetrahedron. The second of these two possibilities can be eliminated on two grounds. In IV, rotation of the RU(CO)~Lgroups would bring the phosphite ligands into trans positions where P-P coupling would occur, as it does in V and VI, and a time-averaged virtual triplet should result, which is not the case. In addition, low-temperature studies of VI in dichlorofluoromethane showed no change in the nmr spectrum on cooling to - 100". This suggests a very low activation energy for the tautomeric process, consonant with exchange of hydrogen atoms,9 parallel to hydrogen exchange in boron hydrides. We believe similar degenerate tautomerism very likely is occurring in the unsubstituted carbonyl hydride I1 and could be ascertained from the 13C satellites of the metal hydrogen resonance; solubility problems, however, have thus far prevented us from observing these. Hydrogen tautomerism in the edgebridged H4Ru4group might well involve face bridging in the transition state. It is interesting to note in this connection that the face-bridged hydrogen atoms of I are easily converted2 to what is most certainly edge bridging in tetrahedral [HBRe4(CO)12]2-.Tautomerism (9) Time-averaged equivalence of ligands has been observed in C4HsFe(CO)(PFs)l and GHsFe(PFa)a at room temperature, but rotational motion was effectively frozen out at about -100"; see J. D. Warren and R. J. Clark, Znorg. Chem., 9,373(1970),and references cited therein. (lo) (a) R. E. Williams, J. Inorg. Nucl. Chem., 20, 198 (1961); (b) W. N.Lipscomb, Science, 153, 377 (1966).

Journal of the American Chemical Society

in I is therefore also possible, as it is in any cluster with unoccupied polyhedral sites lying energetically close to the sites occupied by the bridging hydrogen in the ground-state structure. S. A. R. Knox, H. D. Kaesz* Contribution No. 2839 Department of Chemistry, University of California Los Angeles, California 90024 Received May 24, 1971

Thermochemistry of Carbonyl Addition Reactions. I. Addition of Water and Methanol to Hexafluoroacetone Sir: While the thermochemistry of water and methanol addition to carbonyl compounds has been of significant interest, the studies have been limited to the solution phase' and little thermochemical data are available on the formation of carbonyl hydrates or hemiacetals in other phases where solvation and intermolecular interactions play a smaller role. We wish to report the heat of addition of methanol and water to a ketone in the solution, "standard state," and gas phases at 25". Hexafluoroacetone (HFA) reacts rapidly and completely with water and methanol to form the stable hydrate2 1, mp 43-45" and hemiacetal3 2, bp 95", 0

OH

+

C F J ~ C F ~ROH

e CH3k CFI I

OR

l,R = H

2, R

CH,

respectively. The heats of addition in solution were determined by adding a known quantity of HFA gas t o distilled water or anhydrous methanol in a solution calorimeter described elsewhere. * The heats of solution and vaporization of the hydrate and hemiacetal were determined by direct calorimetry and the data were treated according to the enthalpy cycle. Since HFA(g)

+ ROH(g) eHFA*ROH(g) ?*H"

HFA(g)

.1

+ ROH(1) e HFA. ROH(1 or c)

AH(aa~n)

HFA(so1n)

.1

4H(so1n)

AH,"(,)

t4H7

4

AH,"

4H(~aln)

+ ROH(so1n) eHFA .ROH(soln)

AHro(so~n)

+

HFA is added to the solvent AHeXptl= AHso1n(HFA) SO that AH," = AHexptl AHmln(R0H) AHsoln(adduct)and AH,"(,) = AH," AHv(adduct) AH,(ROH). These expressions can be simplified further since dilute solutions are involved (cn. 4 X M in adduct), AHSoln(ROH)E 0. The experimental quantities are given in Table I. The data clearly show that the heat of hydration and hemiacetal formation of HFA are reasonably close in any one phase. The methanol addition is slightly more exothermic in the "standard state" and slightly less in the gas phase. Presumably, the same trend will obtain AH:(soln)

+

+

(1) (a) R. P. Bell, Aduan. Phys. Org. Chem., 4, 1 (1966); (b) J. Kurz, J . Amer. Chem. Soc., 89, 3524 (1969); J. M. Bell, D. G. Kubler, P. Sartwell, and R. G. Zepp, J . Org. Chem., 30, 4284 (1965). (2) W. J. Middleton and R. V. Lindsey, J . Amer. Chem. Soc., 86,4948 ( 1964). (3) I. L. Knunyanto, T.-Y. Chen, and N. P. Gambaryan, l z v . Akad. Nauk SSSR, Otd. Khim. Nauk, 686 (1960). (4) F. E. Rogers, J. Phys. Chem., 75, 1734 (1971).

1 93:18 September 8, I971

4597 Table I. Enthabies (25". kcal/mol) Adduct (CFa)2C(OH)z (CF&C(OH)OCH3

AHexptl

AHsoln

AHv

AH,"

AHr"(,)

AHr O(so1n)"

-24.4 3Z 0 . 2 -24.7 f 0 . 3

-3.99 3Z 0.04 -3.44 f 0.02

17.0 3Z 0 . 2 17.4 f 0 . 2

-20.4 f 0 . 2 -21.3 f 0 . 3

-13.9 f 0 . 4 -12.8 f 0 . 5

-22.4 f 1 . 2 -22.7 f 1 . 3

~

0

Derived from the relation AH,'

-

AHro(so~n) = AHaoln(HFA)- AHsOin(adduct),assuming AHBoin(HFA)= - 2 f 1 kcal/mol.

with other carbonyl compounds. In both cases the "standard-state" reaction differs only slightly from the heat of reaction in solution. According t o the enthalpy cycle, this latter difference is limited to the difference in the heat of solution of the carbonyl compound and its adduct. The heat of solvation for both reactions, AHro(so~n) - AHro(g),is about -9 kcal/mol. Since it is unlikely that the heat of solvation of the reactants will exceed those of the product, the heat for the solution reaction should represent the maximum or limiting value for this type of reaction. The scarcity of reliable data makes comparison with other gas-phase reactions undesirable; however, comparison can be made with the hydration of similar carbonyl compounds in solution. Table 11 shows that the heat of hydration of HFA is

above, one may then infer that two CFa groups destabilize a carbonyl bond by ca. 12 kcal. This conclusion is in agreement with the remarkable reactivity of HFA7in carbonyl addition reactions. Another method of assessing strain in a molecule is the comparison of experimental heats of formation with calculated heats of formation by a method of group increments.* The heats of formation derived from our experimental data were compared t o values calculated from the group increment scheme of Cox and Pilcher.8b These values are given in Table 111. Table LII. Heats of Formation (25". kcal/mol)

Compd (CFs)zC(OH)r

Table 11. Heats of Hydration of Carbonyl Compounds (kcal/mol)

State

Obsd

Calcd

calcd)

(8)

-370.7

-409.13

38.4

- 359.9 -377.3 -299a

-405.20

45.3

-352.8

53.8

-,

I\ C )

(CFa)zC(OH)OCHs

(8)

(1) CFaC(=O)CFs

-2.0" -5.54 -5.10 - 14.0" -22.4

CHsC(=O)CHzCl CHsC(=O)CHCln CHsCHO CC13CHO CF3C(=O)CFs 5

Reference la.

much larger than analogous compounds and also the heat of reaction increases with halogen substitution (ca. 3-3.5 kcal/halogen atom). Levy, Cargioli, and Racela have reported that multiple halogen substitution also produces an approximately linear decrease in ketone basicity (ca. 2 Ho units per halogen).5 The heat of hydration of acetone is too low to have been measured reliably but assigning AHr(soln)= -1 f 1 kcal/mol allows the formulation of the following equilibrium 0

I1

CFaCCFs

OH

OH

0

+ CH3C-CH3 e CF&CFa + CHIECH3 l I

OH

I

(ObSd Strain-

--AHro-

0

(g)

-3387 .7.

Reference 7.

Table 111 predicts a large amount of strain in HFA which is relieved in the hemiacetal or hydrate by 9 and 15 kcal, respectively. This strain is not apparent in the structure of HFA; the bond angles, carbon-oxygen, and carbon-carbon bond distances of HFAg and acetonelo are the same within the combined experimental errors. (7) C. G. Krespan and W. J. Middleton, Fluorine Chem. Rev., 1, 145 (1967). (8) (a) S . W. Benson, "Thermochemical Kinetics," Wiley, New York, N. Y., 1968; (b) ref 6, p 588. (9) G. Boulet, Dim. Absh., 25, 3283 (1964). (10) C. Kato, S. Konaka, T. Ijima, and M. Kimura, Bull. Chem. SOC. Jap., 42, 2148 (1969).

F. E. Rogers,* R. J. Rapiejko Department of Chemistry, University of Dayton Dayton, Ohio 45409 Received June 17, 1971

OH AH = -21 kcal/mol

The rather large enthalpy associated with the disproportionation reaction can be analyzed in terms of carbonyl group stabilization. In substituted alkenes it is customary to regard differences in heats of hydrogenation as a measure of the stabilizing effect of the substituent. Similarly the heats of hydrogenation of formaldehyde (-22 kcal/mol) and acetone (- 13 kcal/mol)6 show a 9-kcal stabilizing effect for two methyl groups at a carbon-oxygen double bond. To account for the enthalpy in the equilibrium reaction (5) G. C. Levy, J. D. Cargioli, and W. Racela, J . Amer. Chem. SOC., 92, 6238 (1970). (6) Derived from heat of formation data reported in J. D. Cox and G. Pilcher, "Thermochemistry of Organic and Organometallic Compounds," Academic Press, New York, N. Y., 1970.

Transition Metal Promoted Isomerizations of Highly Strained Polycyclic Systems. A Mechanistic Insight' Sir: The rearrangement of highly strained ring systems by transition metal catalysts has recently attracted considerable attention. Although many examples of such rearrangements are now known, relatively little is available in the line of definitive evidence concerning the mechanism of these isomerizations. A typical case is that of tricyclo[4.1 .0.02J]heptane (I), which, (1) Paper XXIV on the Chemistry of Bent Bonds. For the previous paper in this series, see P. G. Gassman, G. R. Meyer, and F. J. Williams, Chem. Commun., in press.

Communications to the Editor